Closed Fs00 closed 2 years ago
whew-inc
commented
3 years ago Hey @Fs00 !
This looks good. I'll test this when I find the time. Thanks for the PR!
As for the lsposed branch/tag, you're right. After testing/merging your proposal I'll clean the repo. Sorry for the confusion.
j0110
commented
2 years ago Is there any updates on this PR ?
whew-inc
commented
2 years ago @Dev0110 I'm currently in a situation irl that makes me unable to work on this or anything really.
If all goes fine expect an update or merge somewhere next week.
whew-inc
commented
2 years ago It looks like it works on Android 11 with EdXposed, but I don't know about Lsposed and I don't want to risk installing right now because I'm lacking time and backups. If someone else can test and confirm it works especially on 11 then I'll merge it, otherwise I'll wait a bit before I do.
Fs00
commented
2 years ago Great! At this point, I think that the odds that this PR won't work on LSPosed/Android 11 are extremely low, given that it works with EdXposed. I've been using FakeGapps with this PR applied on LSPosed on Android 9 and 10 since a few months and everything works fine.
rhjdvsgsgks
commented
2 years ago it works on my device, Android 11 + lsposed
whew-inc
commented
2 years ago Great! At this point, I think that the odds that this PR won't work on LSPosed/Android 11 are extremely low, given that it works with EdXposed. I've been using FakeGapps with this PR applied on LSPosed on Android 9 and 10 since a few months and everything works fine.
I'm only worried about the activation scope not working properly on 11. I tried using the feature with EdXposed but it seems no matter what I select signature spoofing always works
whew-inc
commented
2 years ago It looks like it works on Android 11 with EdXposed, but I don't know about Lsposed and I don't want to risk installing right now because I'm lacking time and backups. If someone else can test and confirm it works especially on 11 then I'll merge it, otherwise I'll wait a bit before I do.
On second thought I'll just merge this now. It works fine on 11 with edxposed and Lsposed should work fine too, just not sure about activation scope. If it doesn't the older apk can be used in the meantime.
Fs00
commented
2 years ago I tried using the feature with EdXposed but it seems no matter what I select signature spoofing always works
It always works because the module has been changed to hook only the Android system server, which corresponds to the "System framework" app when selecting activation scope on LSPosed. I honestly don't remember if you can toggle modules for the "system framework" app in EdXposed.
Anyway, thanks for testing and merging!
whew-inc
commented
2 years ago It's rather strange. Even if I don't select ANYTHING or an irrelevant app, spoofing still works. EdXposed moment? Persists with reboots too.
Fs00
commented
2 years ago Definitely strange. Maybe a bug in the activation scope implementation of EdXposed?
whew-inc
commented
2 years ago Ah, it's because my rom already has MicroG spoofing support, rendering this module redundant for me.
Hello @whew-inc! I did the move to LSPosed, but soon after I found frustrating having to enable FakeGApps for all apps that need GMS signature spoofing. Another consequence of this is that - due to how LSPosed and EdXposed 0.5+ work - there is a small (albeit noticeable) lag on start-up of hooked apps. So in order to fix these inconveniences, I've taken inspiration from LineageOS for MicroG spoofing patches and slightly reworked the logic to only hook into the PackageManagerService class of the system server, precisely after the method generatePackageInfo. This private method has proved to remain stable throughout the years and is present even on very old Android versions (including 4.0). I can confirm this works on Android 6 (original Xposed) and 10 (LSPosed), but should work on other versions too.
I've also added .idea folder to gitignore and bumped app version (feel free to revert/change that if you don't agree).
PS: I've seen that in this fork you have both a branch and a tag named
lsposed. This caused me some trouble when trying to push the code due to this ambiguity. I think it would be best to delete and recreate the tag with a different name.